RETROFIT OF A COLLAPSED MASONRY GALLEY WITH …irf/Proceedings_IRF2016/data/... · 2016-05-27 · Since the XV century studies about the static vaults were developed, but the first

Proceedings of the 5th International Conference on Integrity-Reliability-Failure, Porto/Portugal 24-28 July 2016

Editors J.F. Silva Gomes and S.A. Meguid

Publ. INEGI/FEUP (2016)

-985-

PAPER REF: 6345

RETROFIT OF A COLLAPSED MASONRY GALLEY WITH

HISTORICAL VALUE: PRELIMINARY STUDY OF ALTERNATIVE

SOLUTIONS AND RETROFITTING STRATEGIES

Enrica Di Miceli1(*)

, Vincenzo Bianco1(*)

, Maria Grazia Filetici2, Giorgio Monti

1

1Department of Structural Engineering and Geotechnics (DISG), Sapienza University of Rome, Italy 2Superintendence for Cultural Heritage of Rome, Piazza dei Cinquecento 67, 00185 Rome, Italy (*)Email: [email protected]

ABSTRACT

This work presents a preliminary study aiming at singling out the most suitable design

solution for the retrofitting of a collapsed masonry barrel vault placed in the central

archaeological area of Rome, in Italy. Main objectives to be pursued with the restoration are:

1) re-opening the monument to the public, and 2) endowing it with a walkway on the extrados

of the vault in order to allow visitors enjoy the sight of Circus Maximus. The paper is divided

into four parts, as follows: 1) firstly, a description of the current status of the monument is

provided, then 2) based on the results obtained by simple mechanical models, the possible

causes of the collapse are discussed, 3) some reconstruction strategies are presented and

compared to each other, also in economic terms, and eventually 4) the motivated selection of

the final solution to be implemented is presented.

Keywords: Architectural heritage, barrel vault, monument, preliminary design, retrofitting.

INTRODUCTION

The conservation and restoration of an archaeological monument are a challenge for engineers

and architects.

Most often, these artefacts are the result of a long process of transformations and

stratifications that have occurred over the centuries, and that have ended up modifying the

monument in terms of a) original structural behaviour, b) form and c) function, so that the

result with which technicians have to deal with is a very complex and unique structure.

Uniqueness and complexity require special attention and care both in the evaluation of

techniques and technologies to be used in the restoration (either consolidation or seismic

amelioration design), and in the choice of materials. In fact, when working on an important

monument of historical-architectural value, the design solution to be searched is always the

one that guarantees the best compromise between need to intervene and necessity to bring the

minimum disturbance to both the structure and its surroundings.

Moreover, it is essential to understand which materials to use in order to build any missing

and/or additional parts, without altering the “image” of the artefact. For this reason, it is

necessary a multidisciplinary approach that integrates all different expertise, hopefully

involving architects, archaeologists, historians, and engineers (e.g. Roca et al. 2010).

Stability and durability are also necessary conditions, but not sufficient to define the criteria

and to choose the intervention techniques, having to prefer the preservation of the monument

Symposium_13: Assessment and Strengthening of Existing Structures Subjected to Seismic Demands

-986-

in its environmental context and in accordance with its forms and structural design (DPCM

2011).

The restauration works to be carried on archaeological artefacts must be guided by the criteria

of both “integration” and “integrality” (e.g. Elia 2010).

The present work deals with the study of alternative restoration and seismic amelioration

solutions of a vaulted gallery, in the Central Archaeological Area of Rome. This latter

includes important archaeological monuments and offers a vast range of typologies and

typical (own) Roman architecture construction techniques (e.g. Lugli 1957).

In particular, these monuments are characterized by the presence of arches and vaults, that the

Romans built in various forms and in various ways. Initially, the arches were made of stone

blocks or bricks while later, with the introduction of the pozzolanic mortar, particularly

resistant, and mixed with stones of various sizes, vaults were built by the technique of the

concretion, also called opus caementicium. These vaults were designed as a monolith that is

as a single block of stone that would not have to exert horizontal load on retaining walls.

a)

b)

Fig. 1 - Bricks and vault: a) size of the bricks used in roman constructions; b) three-dimensional view of an opus

caementicium vault with the semilateres ribbing (image from Tomasoni 2015).

In fact, the opus caementicium could not ensure such a monolithical structural behavior since

it does not have a reliable tensile strength and for this reason, the occurrence of cracks turned

the vaults into pushing systems. From a construction point of view, the opus caementicium

was poured on a temporary structure made of wooden ribs that was subsequently removed

Proceedings of the 5th International Conference on Integrity-Reliability-Failure

-987-

after its hardening; for this reason, in structures built by this technique, traces of such wooden

ribs remain impressed in the intrados surface.

Later, Romans began to select the materials constituting the various layers of the vaults and to

sort such materials according to their specific weight, and implementing them according to

the static needs. For instance, in the dome of the Pantheon, light volcanic stones, that are

pumice stones, can be found in the top layers (e.g. Croci 2005). From the I century a. D., the

opus caementicium vaults were built with the so called semilateres that are brick ribbings

positioned along the directrix of the barrel vault and connected to each other with bipedales

bricks (Fig. 1a, b). These ribbings had the function to stiffen the structure, and to confine the

fresh concrete mass.

Any design proposal that involves the statics of a historic building should be based on: 1) an

accurate geometric and material survey, and 2) the interpretation of the static behaviour, so as

to highlight the critical points of the building (yielded from either original conditions,

intentional transformations, degradation or accident).

For this purpose, a part of this work has been dedicated to understanding the causes that may

have led to the collapse of the gallery, by means of simple mechanical models, to verify both

its static and seismic safety. This study is being essential to direct the development phases of

the design strategies and to define what might be considered the best solution in terms of

"advantages and disadvantages", i.e. in terms of a) total or partial achievement of all the goals,

b) respect of the restrictions, and c) the level of operational and economic feasibility.

Fig. 2 - Urban localization: general view of the central archaeological area of Rome.

Since the XV century studies about the static vaults were developed, but the first static

theories about the arches, vaults and domes, date back to the XVII century. Important is, for

example, the de La Hire’s contribution (e.g. de La Hire 1730,1731) who first used the

funicular polygon construction in the static analysis of the arches. Couplet (e.g. Couplet

1731,1732) introduced the concept of rotational hinge as a condition for the fracture.

Coulomb (e.g. Coulomb 1776) based his studies on finding the collapse safety factor and

introduced the friction and cohesion conditions between the blocks. During the XIX century,

starting from Navier, attention was also paid to the strength of materials to determine the

application points of resultant forces in the keystone and in the fractures joints. An early study

about the pressure curves was conducted by Mery (e.g. Mery 1840) who proposed a graphical

verification method of the arches that is still used nowadays.


-988-

Currently the analysis and verification methods about the safety of arches and vaults are

numerous. The most significant studies developed over the past years encompass: 1) the

theory of plasticity and limit analysis applied to masonry structures (e.g. Heyman 1996, 1982;

Monti et al. 2013; Tomasoni 2015); 2) the membrane theory, which takes into account the

three-dimensional effects (e.g. Flugge 1973); 3) the analysis of the vaults through the finite

element modeling according to a "continuous" or a "discrete” approach (e.g. Lourenço 2002;

Roca et al. 2010). In particular, the limit analysis method has been proven to be one of the

most powerful techniques for engineering practice, due to the suitable compromise between

relative simplicity and prediction accuracy. In a preliminary phase, this method appears the

most appropriate and effective tool to verify the structures.

CURRENT STATUS OF THE MONUMENT

The so-called Galleria delle Volte Crollate (The Collapsed Vault’s Gallery), object of this

work, is part of an important archaeological site – the Augusto’s House on the Palatine hill –

which has not been completely explored yet. This site is therefore the subject of preliminary

assessments, both historical and archaeological, which are being studied by experts in this

field.

Fig. 3 - Previous retrofit intervention (XX century): plan with optical cones of images: details of the steel beams,

squat brick walls and wooden struts.

The Gallery is located in the central archaeological area of Rome, precisely in the south-west

area of the Augusto’s House (Fig. 2) and it was originally roofed by a depressed barrel vault

made by a conglomerate composed of irregularly shaped tuff stones and pozzolanic mortar,

according to a technique called opus caementicium. At a certain point of the history, and for

unknown reasons, the vault must have collapsed disintegrating in large conglomerate blocks.


-989-

These resulting blocks, probably remained buried under soil and rubble for a long time, were

rearranged, according to a restoration design of the second half of the twentieth century

(between 1960 and 1980), by a complex and heterogeneous system comprising 1) steel

beams, 2) squat brick masonry walls, and 3) wood struts (Fig. 3). That design was meant to

restore the vaulted passage in order to make it usable by visitors. Only some portions of the

vault are still standing in their original place. The structures beneath the quota in question,

including the foundations, have not been excavated yet.

Anyway, the gallery is currently closed to the public. The status of the gallery, at a first sight,

does not appear to be safe enough to allow access to the visitors. Among the aspects that give

rise to some concern, are: 1) some of the wood struts are cracked, slightly damaged at their

extremities and probably not safe with respect to the buckling phenomenon, 2) some of the

steel beams are oxidized and with a flexural curvature that can be appreciated by naked eye

(Fig. 3), 3) static safety in some points relies on mutual contrast and interlock between

adjacent blocks that, in correspondence of the contact point, due to the high concentration of

stresses, are visibly cracked with resulting dusty rubble and likely loss of stability. Moreover,

in correspondence of the north-western entrance to the gallery, the downhill wall is

undergoing an outward rotation.

An accurate survey of the monument was carried out in order to determine 1) the geometrical

dimensions and relationships among the different parts it is composed of, 2) the constitutive

materials, their state of conservation, and relevant physical and mechanical properties

(qualitatively, by comparison with literature data), 3) the original structural conception and

possible successive modifications. As to the geometrical features, the vault is a barrel

depressed arch and its spring line (L) is 4.45m and its rise (f) is 0.59 m (Fig. 7). For what

concerns the materials: 1) the “structural part” of the vault is made of tuff conglomerate and

pozzolanic mortar; 2) the filling layer is constituted by two layers: the former in mortar and

flint aggregates and the latter, a thinner layer, in “cocciopesto” aggregates (Fig. 4). The

extrados shows traces of the so-called “suspensurae” (brick walls spaced apart that created a

ventilated interspace) and a pavement in bessali (Figs. 1a, 4) that leaves guess that the gallery

was also accessible at the extrados (Fig. 4).

Fig. 4 - The constitutive components of the Gallery vault.


-990-

The gallery walls are characterized by different construction techniques. On one side, the

vault lied on a brickwork masonry (opus testaceum) and on the other, on a mighty wall of tuff

and travertine blocks that today stands at a height much lower than the original one.

A detailed survey of all the steel beams was carried out (type, span, degradation), by

numbering 41 beams in total (39 double T steel beams and two wooden ones). The safety

verification about the steel structures (Di Miceli 2016) has highlighted a situation of possible

risk: a) 29 beams are not verified for flexural static loads, and b) 4 are not verified for both

flexure and shear (Table 1). These first results, therefore, suggest to reconsider the twentieth -

century intervention, eliminating the steel and wood supporting structures and proposing

possible alternative solutions for the vaulted roof of the gallery, according to a more correct

interpretation of the monument structures.

Table 1 - Excerpt of the results of the steel beams safety analysis (Di Miceli 2016).

POSSIBLE EVOLUTION OF THE MONUMENT

Starting from the study of the bibliographic sources (e.g. Iacopi et al. 2006; Pensabene et al.

2011, 2013; Carandini et al. 2010), two different hypotheses about the architectural genesis of

the so-called Galleria delle volte crollate were taken into account:

1) the Gallery may be considered a vaulted substructure lying on the soil, and directly

cast on it, as a support of the bases in front of the Apollo’s Temple, which was part of the

Augustus’ House (Fig. 5). From a structural point of view, therefore, its stability would be

related to the bearing capacity of the underlying soil;

2) the Gallery is a self-supporting vaulted structure, with depressed cross-section, and its

stability is related to the mechanical properties of materials and its geometric shape.

The first hypothesis would be supported by the fact that, when Augustus decided not to futher

expand his house in favor of the construction of the Apollo’s Temple, he also decided to bury

all the previous structures and to raise the walking level. The construction of the vaulted


-991-

gallery would be, therefore, contemporary with the construction of all the substructures in

front of the Apollo’s Temple. Another element in favor of the first thesis is that the barrel

vault would result supported, along its longitudinal development, on vertical structures made

in different ages and by different techniques, that are: 1) on the southern side, on the remains

of a brickwork masonry and, 2) on the northern side, on a wall with tuff and travertine blocks

(Fig. 5).

In support of the second hypothesis, however, some literature sources claim that the Gallery

would have been built as a vaulted walkway that connected the rooms of the western peristyle

of Augustus’ House with the structures of the eastern peristyle (Fig. 5). These latter have a

vault that is very similar, in terms of both materials and geometry, to that of the Gallery in

question and therefore it was very probably built in the same timeframe (Fig. 6a). Another

element, which would suggest that the gallery was built as a self-supporting structure, is

represented by the traces of the wooden planks still recognizable at its intrados (Fig. 6b).

Fig. 5 - General plan of the Augustus complex.

In the next section, all these assumptions have been reevaluated also on the basis of the results

obtained by means of simple, yet efficient computing models, trying to identify the causes

that may have led to the collapse of the vault, thus trying to answer the following questions:

1) Did the vault collapse due to the fact that it actually was a depressed vault?

2) Did the weight of the fill above, not foreseen by the original design, cause the collapse

either in static or seismic conditions?

3) Was the collapse of the vault caused by the illegal removal, according to an unfortunate

and widespread practice of the time, of some tuff and travertine blocks of the perimeter walls

thus weakening the supports of the vault itself? Note that similar episodes have been

unfortunately occurring in other monuments of the Palatine.


-992-

a)

b)

Fig. 6 - Similar vault to that of our Gallery (a); traces of the wooden planks at the intrados of the vault (b).

PROBABLE CAUSES OF THE COLLAPSE

The safety of the arch, in its original features before collapse, was verified by means of simple

mechanical models in order to have an expeditious and preliminary estimate. The arch safety

was analysed against both static and seismic loads. Such a study can also be useful 1) to avoid

to reconstruct a structure that was affected by design mistakes and 2) to better address the

decisions to be taken in order to retrofit the gallery and make it usable by the visitors of the

Palatinum Area.

Fig. 7 - Scheme of the vault (a), and discretization in blocks of both the filling and the structural arch (b).


-993-

Static Loads

The verification under static loads was carried out by means of Mery’s Method (Mery 1840),

verifying if, for a uniformly distributed load (self-weigh plus fill), the thrust line is contained

within the core of the arch. In that case, the arch would result safe, according to Mery.

A one-meter deep portion of the barrel vault was analysed. The Mery’s method was translated

into analytical formulations and implemented in a simple computer program. The arch was

discretized into a large number of blocks in order to calculate the various involved

parameters, that are, for instance: the weight of the vault and its overlying fill (Fig. 7).

The verification shows that the thrust line is entirely contained within the core (Fig. 8a). But

this result is not reliable enough to draw positive conclusions about the safety of the arch

since the model is affected by a drawback. In fact, since it imposes the passage of the force

resultant through both 1) the uppermost extremity of the core, in correspondence of the

keystone, and 2) the lowermost extremity of the core, in correspondence of the imposts, it

ends up providing, for static distributed loads, and whatever geometrical contour conditions, a

thrust line that is always contained in the arch core. Thus, that result could be misleading.

So a normal stress verification was carried out in correspondence of the arch sections in

which the vertical weights, due to both the filling and the blocks’ self-weight, are combined

with the horizontal force applied to the keystone (Figs. 8b and 9).

Fig. 8 - Static analysis: trust line (a); trend of the ratio between compressive masonry strength and demand (b).

For the mechanical properties of the constitutive material, rubble tuff block masonry,

(Circolare Esplicativa n°617) the unit compressive strength was assumed equal to 2

0.140 cmNR =σ . The trend of the ratio between the masonry strength and demand, along

the arch length, in terms of compression SR σσ , is plotted in Fig. 8b. As can be gathered

from this trend, which is plotted for different values of the structural masonry unit weight (30.250.15 mkNm ≤≤ γ ), the ratio SR σσ ranges from a minimum value of 50.2 and a

maximum value of 20.7 . Note that the trend of the ratio SR σσ along the arch shows a

singular point with a change of curvature approaching the imposts that is due to the change of

the sign of the thrust line eccentricity with respect to the arch mid line. It emerges that the

arch was properly designed to support the gravity loads to which it was subject.


-994-

Fig. 9 - Verification of the sections: scheme of the normal component (N) along the transversal section.

Seismic Loads

The verification under seismic loads was carried out by means of the Limit Equilibrium

Analysis Method, assuming for the arch a collapse mechanism according to which four hinges

divide it into three rigid blocks (I, II and III in Fig. 10). Such labile mechanism has one

degree of freedom that is the downwards rotation 1θ of the block 21CC , while the rotations of

the remaining two blocks, 2θ and 3θ around their relevant instantaneous rotation center are

kinematically dependent on 1θ . The collapse mechanism, due to a seismic action directed

rightwards is represented in Fig. 10. In searching for the most probable collapse

configuration, coherent with the collapse mechanism above, the following assumptions were

made: 1) only the case of arch portions contained in an opening angle larger or equal to °15

were taken into consideration, 2) the first hinge 1C was assumed to be located at a distance of

about %5 of the arch springing line (L) from the left impost, and 3) the fourth hinge 4C was

assumed to be located at the right impost.

For the properties of masonry, the following assumptions have been taken: a) no tensile

strength; b) infinite compressive strength; c) absence of any slippage, i.e. infinite friction,

along the interfaces between the tuff irregular blocks and the mortar.

The horizontal seismic force that activates the overturning mechanism is defined by means of

the horizontal loads multiplier 0α , which is the mechanical ratio of the work done by the

stabilizing forces with respect to the work done by the overturning forces, in accordance to

the Virtual Work Principle (Circolare Esplicativa n°617):

�� ∙ �� ∙ ��,� � � ��

�� ∙ ��,�� ∙ ��,� � �� (1)

stabovrt LL =⋅0α (2)

where: iP is the self-weight of the i-th rigid block, jP is the j-th carried weight, izd , is the

virtual horizontal displacement of the point of application of iP , jzd , is the virtual horizontal

displacement of the point of application of jP , iyd , is the virtual vertical displacement of the


-995-

point of application of iP , and fiL is the work of internal forces. Note that when the vertical

displacements are directed downwards they positively contribute to the overturning work (

ovrtL ), and vice-versa ( stabL ).

The value Ra of the spectral seismic acceleration can be evaluated as follows (Circolare

Esplicativa N°617):

�� ∙ ∑ ��∗ ∙ �� ∙ ��∗ ∙ �� (3)

where CF is the Confidence Factor, function of the Knowledge Level KL ; g is the gravity

acceleration; and *M is the participating mass, which can be evaluated by (Circolare

Esplicativa N°617):

�∗ � ∑ �� ∙ ��,�� !"� ∙ ∑ �� ∙ ��,�" (4)

The demand acceleration Da was assumed as the value corresponding to the constant

acceleration plateau of the site spectrum (Fig. 11 and Table 2), evaluated by means of the

Italian Standards for Constructions (NTC2008), and four limit states, that are: 1) Immediate

Occupancy Limit State (IOLS with return period yrsTR 30= ), 2) Damage Limit State (DLS

yrsTR 50= ), 3) Life Safety Limit State (LSLS yrsTR 475= ), and 4) Collapse Limit State

(CLS yrsTR 975= ). The capacity/demand ratio was calculated as follows:

D

R

a

a=ρ

(5)

Fig. 10 - Seismic analysis: assumed collapse mechanism.

In searching for the minimum value of the collapse multiplier, all the possible collapse

configurations consistent with the assumptions above, were analyzed. In doing that, the ratio


-996-

between the positive amount of the stabilizing work stabL, in case of vertical displacement

with direction opposite to the relevant weight load ()()( , iiy Psigndsign ≠), and the negative

amount of the stabilizing work, in case of vertical displacement with the same direction as the

relevant weight load ()()( , iiy Psigndsign =), varies almost with continuity. In this way, the

collapse multiplier can also assume either negative or very low positive values. The former

ones were discarded since they do not have physical meaning while the amount of positive

values were analyzed in statistical terms (Fig.12). In fact, assuming the minimum value of 0α

would result excessively conservative.

Fig. 11 - Comparison between response spectra.

Table2 - Seismic parameters

ag [g] Fo Tc*[s] SS CC ST

IOLS 0.043 2.539 0.256 1 1 1.4

DLS 0.055 2.502 0.269 1 1 1.4

LSLS 0.121 2.630 0.293 1 1 1.4

CLS 0.152 2.615 0.301 1 1 1.4

For each of the analyzed limit states, a normal distribution of the probability density was

determined and the relevant value of the probability of attainment of the LS, conditioned to

the given value of the peak ground acceleration, was calculated as follows:

( )[ ] ρρ dDLSag ⋅=≤ ∫1

0

Pr|0.1Pr

(6)

where: PrD is the probability density function. The results are plotted in Fig. 12. As can be

gathered from that figure, the probability of failure of the several LS is: 18% for the IOLS,

25% for the DLS, 74% for the LSLS, and 86% for the CLS. This yields that the arch, in its

original shape and loading condition, was not safe with respect to the seismic loads given by

the Italian Regulation currently in force (NTC2008). So, it is possible to conclude that the

arch might have collapsed due to an earthquake attack.


-997-

The results of the calculations would support the hypothesis that the gallery’s vault was

conceived as a structural one thus disproving the hypothesis that it was obtained by simply

casting the opus caementicium directly on the soil. Moreover, the traces of the wooden

planks, still visible on the intrados of the vaults remaining portions, would also support the

second hypothesis according to which the gallery was a structural one.

Fig. 12 - Probability density of the seismic capacity/demand ratio for different limit states:

Immediate Occupancy (IOSL), Damage (DSL), Life Safety (LSLS), Collapse (CSL).

In light of the considerations above, we can conclude that the vault, even though intentionally

conceived as a structural one, may have been affected by macroscopic design defects so that it

was not able to withstand seismic loads and probably collapsed due to an earthquake.

ALTERNATIVE RETROFITTING STRATEGIES

In this section, several restoration design alternatives are evaluated in terms of both feasibility

and economical expense. For each of such hypotheses, all the construction phases were first

singled out and subsequently evaluated in economic terms.

Fig. 13 - Current state of the Gallery (a); integration of perimeter walls and cutting of steel beams flush against

the wall so that the left portions remain incorporated in the new masonry (b).


-998-

The main objectives to be pursued are: 1) making the gallery accessible to the public, 2)

recreating the original intrados surface, 3) increasing the safety of the structure for both static

and seismic loads, 4) suitably organize the construction site, given the delicate surroundings,

and 5) bring the minimum disturbance to both the supporting lateral walls and the underlying

structures, down to the foundations. In fact, as to this latter point, the structures beneath the

walls still have to be excavated and are partially unknown.

The cost analysis was carried out indicating for each constructive phase the relevant a) unit of

measure (u. m.), b) unit price (u. p.), and c) quantity of each specific component, in relation to

one linear meter of the Gallery’s longitudinal extension (37,27m long by 4,45m wide). Works

that are common to each of the design proposals have been listed in the relevant tables but not

quantified in economic terms, in order to carry out a comparative economic analysis

highlighting only the works peculiar of each solution. Prices were deducted from both

“Tariffa dei prezzi 2012 - Regione Lazio” (TPRL 2012) and “Listino noleggio Pernicini

2010” (LNP 2010).

Fig. 14 - Installation of possible tie rods (a), view after the demolition of the steel beams and the brick walls;

similar tie roads found in another monument in the archeological area of Rome (b).


-999-

Fig. 15 - First solution: bridge crane scheme (a); reconstruction of the vault and the floor above (b).

First solution

1) Masonry reconstruction of the barrel vault with localized lifting and

repositioning of the blocks with an overhead travelling bridge crane

(Figs. from 13 to 15)

u. m.

u. p.

quantity price.

[€/ml]

1.1) The previous intervention in steel beams, wooden supports and brick

walls is initially left in place.

1.2) Integration of the perimeter walls (tuff, brick and travertine) of the

gallery, up to the maximum height of the vault extrados: this height is

deduced from two points in which the vault is still in the original

quota (Fig. 13a,b). If necessary, the perimeter walls are shaped to

create the impost of the vault.

On the north-west side (masonry blocks of tuff and travertine), where

the longitudinal access openings to the adjacent Domus Augusto’s

rooms can be found, the flat arches above such openings will be

rebuilt, at their original impost quota.

1.3) It will be necessary to verify that, after removal of a) wooden

supports, b) squat brick walls or other blocks, the vault portion left in

place is able to support both itself and the steel beams of the deck to

be positioned successively; otherwise it is necessary to intervene

locally with tie rods and consolidating injections (Fig. 14a, b).

ml

22,26

10

32,80

1.4) The collapsed vault blocks, currently stored inside the gallery, but

that are difficult to relocate properly in their original position, due to

the shortage of information, are lifted by the crane and taken to the

external storage area.

month

1.470,00

12

473,30

1.5) Installation of the longitudinal (assuming HEA1000 = 272kg/m) and

transversals beams (HEA 500 = 155kg/m) of the steel deck

(included the bracings), by means of the crane (Fig. 14b).

kg

kg

3,35

3,35

544

155 1.822,40

2.310,66


-1000-

1.6) Bridge crane construction (installation) with two orders, where its

dimensions are calibrated on the basis of both the maximum and

minimum block size (Fig. 15a). Note that the cost includes the

electric winch.

month 14.000,0 12 4.507,64

1.7) The vault blocks are lifted and repositioned, with an electric winch;

construction of the wooden plank arch-centering; these operations

are done by subdividing the gallery length in successive sub-portions.

After completing a sub-portion the bridge crane is moved and

operations are iterated at the adjacent sub-portion.

mq

48,35

4,45

215,15

1.8) Removal of the elements composing the previous intervention (XX

century). The steel beams are cut along the cross section in

correspondence of the wall surface so that the left portion will remain

embedded in the wall and visible to the outside, thus leaving

historical memory of the previous intervention (Fig. 13b); the brick

walls are demolished; transport of end products in landfill.

1.9) Reconstruction of the missing “structural” portion of the vault. mq 230,00 2,225 511,75

1.10) Possible fill in lightweight material (only if it has a static function) to

the extrados of the new reconstructed parts.

1.11) The wooden plank arch-centering is disassembled.

1.12) Construction of a floor with corrugated steel sheet and reinforced

concrete slab: note that the beams depth is such as to leave a space

that allows to preserve the suspensurae (Fig. 15b).

mq

49,69

4,45 221,12

[€/ml] 10.604,06

[€] 395.213,3

Fig. 16 - Second solution: masonry reconstruction of barrel vault.

Second solution

2. Masonry reconstruction of barrel vault with lifting and

handling of blocks by crane (i.e. without bridge crane)

(Figs. 13,16)

u. m.

u. p.

quantity price.

[€/ml]

2.1) Numeration and accurate survey of the current position of the

collapsed vault blocks.

2.2) Consolidation of the vault blocks (for example with

injections) to improve their mechanical properties and to

ensure that – during the lifting stages – they will not undergo

any damage.

2.1) The vault blocks are lifted by crane, positioned outside the

Gallery, and stored in a nearby outdoor area.

month 1.470,00 12

473,30

2.2) Integration of the perimeter walls (tuff, brick and travertine)


-1001-

of the gallery, up to the maximum height of the vault extrados:

this height is obtained by two points in which the vault is still

in the origin quota (Fig. 13a,b). If necessary, the perimeter

walls are shaped to create the impost of the vault.

On the north-western side (masonry blocks of tuff and

travertine), where the longitudinal access openings to the

adjacent Domus Augusto’s rooms can be found, the flat arches

above such openings will be rebuilt, at their original impost

quota.

2.3) Removal of the elements composing the previous

intervention (XX century). The steel beams are cut along the

cross section in correspondence of the wall surface so that

the left portion will remain embedded in the wall and visible

to the outside, thus leaving historical memory of the previous

intervention (Fig. 13b); the brick walls are demolished;

transport of end products in landfill.

2.4) Construction of the arch-centering along the Gallery. mq 48,35 4,45 215,15

2.5) Lifting and correct re-positioning of the vault blocks by the

crane.

2.6) Reconstruction of the missing structural vault portion. mq 230,00 4,45 1.023,50

2.7) Filling in lightweight material to the extrados of the new

reconstructed parts;

cast concrete (4 cm) with welded steel net (mesh 10x10

cm) and construction of a flooring, leaving in high or low

relief the parts of vault with suspensurae (Fig. 16).

mc

mq

mq

132,00

25,00

35,00

2,79

2,225

2,225

368,28

55,62

77,87

[€/ml] 2.213,72

[€] 82.505,3

Fig. 17 - Third solution: bridge crane scheme (a); reconstruction of vault


-1002-

Fig. 18 - Fourth solution: steel bridge-like deck seismically isolated.

Third solution

3) Repositioning of the collapsed vault blocks, integration of the missing

parts with new masonry and construction of a steel bridge-like deck

(not isolated) at the extrados of the barrel vault to which the

underlying structures are eventually hung (Figs. 13, 17)

u. m.

u. p.

quantity price.

[€/ml]



deduced from two points in which the vault is still in the original

quota (Fig. 13a,b). If necessary, the perimeter walls are shaped to

create the impost of the vault.

On the north-western side (masonry blocks of tuff and travertine),

where the longitudinal access openings to the adjacent Domus

Augusto’s rooms can be found, the flat arches above such openings

will be rebuilt, at their original impost quota.

3.2) Consolidation of the vault blocks (for example with injections) to

improve their mechanical properties and to ensure that – during the

lifting and handling stages - they will not undergo any damage.


transversal beams (HEA 500 = 155kg/m) of the steel deck (including

the bracings), by means of the crane (Fig. 17b).

kg

kg

3,35

3,35

544

155 1.822,40

2.310,66

3.4) Assemblage of the bridge crane and its positioning by crane

(Fig.17a).

month

month

14.000,0

1.470,0

12

12 4.507,64

473,30

3.5) Performing the following operations, proceeding by "modules":

3.5.1) lifting and positioning of the first block of the current module;

3.5.2) the block is fixed by tie rods to the transversal beams of the steel

bridge-like deck (Fig. 17a);

3.5.3) the bridge crane is moved on the possible second block of the same

module;

ml

31,00

5,05

156,00


-1003-

3.5.4) lifting, positioning and fixing of the second block (if it exists) of the

current module;

3.5.5) the steel beams of the previous intervention are removed and the

relevant brick walls demolished;

3.5.6) construction of the arch-centering;

3.5.7) predisposition of the tie rods that will sustain the parts to be rebuilt;

3.5.8) reconstruction of the missing “structural” vault portion;

3.5.9) the arch-centering is disassembled;

3.5.10) installation of the tie rods for the parts to be rebuilt;

3.5.11) moving to the next “module”.

mq

mq

ml

48,35

230,00

31,00

4,45

2,225

3

212,15

511,75

93,00

3.6) The bridge crane is disassembled, after every block has been correctly

placed.



that allows to preserve the suspensurae (Fig. 17b)

Mq

mq

49,69

35,00

4,45

4,45 221,12

155,75

[€/ml] 10.461,26

[€] 389.891,1

Fourth solution

4) Repositioning of the collapsed vault blocks, integration of the missing

parts with new masonry and construction of a steel bridge-like deck

(isolated) at the extrados of the barrel vault to which the underlying

structures are eventually hung (Figs. 13, 18).

u. m. u. p.

quantity price.

[€/ml]



obtained by two points in which the vault is still in the original quota

(Fig. 13a,b). If necessary, the perimeter walls are shaped to create the

impost of the vault.



Augusto’s rooms can be found, the flat arches above such openings

will be rebuilt, at their original impost quota.

4.2) Definition and identification of the supports (at least 4) of the deck on

which to place the seismic isolators (Fig. 18a).

4.3) Installation of seismic isolators. piece 2000 2 107,32



lifting stages - they will not undergo any damage.

kg

kg

3,35

3,35

544

155 1.822,40

2.308,15


transversals beams (HEA 500 = 155kg/m) of the steel deck

(including the bracings), by means of the crane.

kg

kg

3,35

3,35

544

155 1.822,40

2.310,66

4.6) Assemblage of the bridge crane and its positioning by crane. month

month

14.000,0

1.470,0

12

12 4.507,64

473,30

4.7) Performing the following operations, proceeding by "modules":

4.6.1) lifting and positioning of the first block of the current module;

4.6.2) the block is fixed by tie rods to the transversal beams of the steel

bridge (Fig. 18b);

4.6.3) the bridge crane is moved on the possible second block of the same

module;

4.6.4) lifting, positioning and fixing of the second block (if it exists) in the

current module;

4.6.5) the steel beams of the previous intervention are removed and the

relevant brick walls demolished;

4.6.6) construction of the arch-centering;

4.6.7) reconstruction of the missing “structural” vault portion (the pour);

4.6.8) the arch-centering is disassembled;

ml

mq

mq

ml

31,00

48,35

230,00

31,00

5,05

4,45

2,225

3

156,00

212,15

511,75

93,00


-1004-

4.6.9) installation of the tie rods for the parts to be rebuilt;

4.6.10) moving to the next “module”.

4.8) Construction of the necessary joints to accommodate the seismic

displacements (Fig. 18a, b).



that allows to preserve the suspensurae (Fig. 18b)

mq

mq

49,69

35,00

4,45

4,45 221,12

155,75

[€/ml] 14.461,26

[€] 538.971,1

Fifth solution

5) Construction of a bridge-like steel truss deck, by means of prefabricated

modules, whose cross-section is such as to recreate the same original

intrados profile. Subsequent processing of the blocks with cutting of the

cortical parts, both 1) at the intrados (curvilinear), and 2) intrados

(flat) in portions to be fastened at the steel truss deck, at the this

latter’s intrados and extrados, respectively. (Those portions could be of

dimensions such as to be handled by naked hand, that is without

crane).

u. m.

u. p.

quantity price.

[€/ml]

5.1) Numeration and accurate survey of the current position of the collapsed

vault blocks.



deduced from two points in which the vault is still in the original quota

(Fig. 13a,b). If necessary, the perimeter walls are shaped to create the

impost of the vault.



Augusto’s rooms can be found, the flat arches above such openings will

be rebuilt, at their original impost quota.



subsequent lifting and handling stages – they will not undergo any

damage.

5.4) The vault blocks are lifted by crane, positioned outside of the Gallery,

and stored in a nearby outdoor area.

month 1.470,0 12 473,30

5.5) Cutting of the cortical parts of the vault blocks by grinding machine:

1) at the intrados (curvilinear portions of the structural arch), and 2) at

the extrados (supported filling including the so-called “suspensurae”)

(Fig. 19).

day 30,00 240 193,18

5.6) Transportation to the construction site of all the modules of the steel

bridge-like truss deck.

per

trip

150,00 8 32,19

5.7) Positioning in place by crane and assemblage of the steel truss modules

(assumed indicatively 1m deep). Steel truss module (U140 = 16kg/m;

L20x20 = 1,14kg/m; U100 = 10,60kg/m) (Fig. 19).

kg

kg

kg

3,35

3,35

3,35

659,20

30,16

473,73

2.208,32

101,03

1.586,99

5.8) Removal of the elements composing the previous intervention. The

steel beams are cut flush against the wall surface so that the left part of

these will remain "incorporated" in the masonry (historical memory of

the previous intervention) (Fig. 13b); transportation of end products in

landfill.

5.9) Positioning in place and fastening to the steel truss deck of the cortical

elements (both intrados and curvilinear and extrados and flat). Such

operations could be carried out, at the intrados, by naked hands, per

small portions while, at the extrados, per larger portions, handled by

the crane. Cortical extrados portions are positioned in such a way to


-1005-

result in high-relief with respect to the surrounding practicable floor.

5.10) Construction of an accessible floor by corrugated steel sheet and

reinforced concrete slab; construction of the practicable floor (in bas-

relief).

mq

mq

49,69

35,00

4,45

4,45

221,12

155,75

[€/ml] 4.971,88

[€] 185.301,9

Fig. 19 - Fifth solution: steel truss deck, with fastened cortical portion of the original barrel vault.

SELECTION OF THE FINAL SOLUTION

In the previous section, all the contemplated retrofit strategies were described in detail.

Advantages and disadvantages of those solutions are discussed hereinafter.

The first hypothesis substantially envisages the following phases: 1) positioning of a bridge

crane, 2) repositioning of the blocks, 3) building the wood plank arch-centering, 4) building

(casting) of the vault, 4) construction of the steel deck at the vault extrados. This solution

presents some aspects difficult to address properly: 1) after building the wooden plank arch-

centering, since the bridge crane should be still in place, it would heavily reduce the freedom

of movement at the extrados during the construction of the new portions of the vault, 2) the

resulting structure would be composed of old and new parts whose interaction would be

extremely delicate and difficult to handle, from a safety calculations standpoint, due to the

inevitable difference in mechanical properties, 3) the expeditious calculations herein

presented showed that the vault was ill conceived so that it is not reasonable to re-built,

faithfully to its original features, a structure that was already affected by huge defects, 4) it is

very expensive due to the hiring of both crane and bridge crane.

The second hypothesis aims at re-building the vault reducing the local operations by 1)

initially moving the blocks outside the gallery perimeter by crane, 2) building the wood plank

arch-centering, 3) building the missing parts of the gallery. This solution, even though less

expensive than the first one, due to the absence of the bridge crane hiring, still maintains the

safety-related issues. In fact, we would realize a structure that was evidently deficient, by also

pretending to rely on the interaction between old masonry and new masonry. This is even

more pretentious if we remark the heterogeneity of the masonry material itself, even more if it

is ancient. With respect to the previous one, this solution presents, anyway, the advantage of


-1006-

reduced costs. Even if, in the present work, for the sake of brevity, the interventions necessary

to make both the vault and its supporting walls earthquake-resistant were not accounted for.

The third hypothesis envisages: 1) the positioning of the bridge crane, 2) building of the wood

plank arch-centering, 3) installation of the steel braces that will support the vault, 4)

realization of the missing parts of the vault that will remain hung at the steel deck realized at

the vault extrados. This solution eliminates the issues related to the interaction between old

and new masonry since the vault would remain hung at the steel deck, but maintains the

feasibility problems related to the interaction of constructive phases that affect the first

solution. Among the advantages of this solution there is also the reversibility (Brandi 1977).

In fact, since the new parts are hung at the steel deck, they could be removed.

The fourth solution envisages the seismic isolation of steel deck to which the vault should be

hung by steel rods. This solution is unfeasible for several reasons: 1) the steel deck would

need point-wise supports (four) with high local stress concentration that would require heavy

interventions on the supporting lateral walls, and 2) large portions of the lateral walls should

be cut in order to accommodate the lateral displacements that the steel deck would undergo

during an earthquake. Such solution would result too invasive, and not in agreement with the

restoration philosophy currently accepted. In fact, the full seismic rehabilitation strategy,

allowed by the 1981 Ministerial Decree (D.M.LL.PP 1981), has been progressively

abandoned in favor of the less invasive seismic amelioration strategy, which is the

intervention design solution currently accepted in Italy for the monuments (LINEE GUIDA

2010).

The fifth solution envisages the following phases: 1) moving the blocks outside of the gallery

perimeter, 2) positioning of the prefabricated steel truss modules on the longitudinal

supporting walls and their assemblage on site, 3) cutting of the cortical part of the blocks,

both intrados (curvilinear) and extrados (flat), 4) positioning and fastening of the cortical

parts, whose dimensions may be such as to allow bare hands movement, 5) construction of the

practicable floor at the extrados. Many are the advantages that seem to characterize this

solution, among which: 1) reversibility, 2) minimum disturbance to the supporting walls, due

to the continuous support, as opposed to the point-wise one of the fourth solution, for

instance, 3) relative inexpensiveness, due to the possibility of prefabrication of the deck in

modules, 4) improving safety by also connecting the two longitudinal walls thus improving

their behavior against seismic actions orthogonal to their longitudinal development.

Moreover, other materials, lighter and less vulnerable to corrosion than steel could be

adopted, such as Aluminum or Fiber Reinforce Polymers (FRP), respectively.

CONCLUSIONS

This paper presented a preliminary study of possible alternative retrofitting strategies of a

collapsed vault placed in the central archeological area of Rome, in Italy. It is a Gallery

placed inside the famous Augusto’s House area. For not documented reasons, this gallery

must have undergone a collapse, in an unknown time of history. Subsequently, in the second

half of the twentieth century, it was retrofitted by a complex intervention based on the use of

wooden struts, steel beams, and squat masonry blocks. This intervention, however, presents

many limits.

By means of simple mechanical models, the safety of the vault, considered in its original

features, showed that it was characterized by design defect that made it vulnerable to seismic

loads. Thus, it may have collapsed due to an earthquake.


-1007-

The fifth, among the retrofitting design hypotheses taken into consideration, resulted the most

suitable one, in terms of a) reversibility, b) expensiveness, c) disturbance brought to the

existing surrounding structures, d) safety. This solution envisages 1) the construction of a

steel truss deck, whose shape aims at reproducing the same intrados profile as the original

gallery, continuously supported along the lateral walls, and 2) the fastening of the cortical

portions only, of the collapsed blocks, to the truss. This solution, also reduces the seismic

vulnerability of both vault itself and supporting walls, by dramatically reducing the masses

involved.

REFERENCES

[1]-Brandi, C., (1977). “Teoria del Restauro”, Einaudi, Torino, 1977.

[2]-Carandini, A., Bruno, D., (2010). “La casa di Augusto, dai Lupercalia al Natale”, Laterza,

Bari 2010.

[3]-Circolare 2 febbraio 2009, n. 617. “Istruzioni per l’applicazione delle Nuove norme

tecniche per le costruzioni, di cui al D.M. 14 gennaio 2008”.

[4]-Coulomb, C., A., de, (1776). “Essai sur une application de maximis et minimis à quelques

de statique, relatives à l’Architecture”. Mémories del Mathématique et de Physique presents à

l’Académie Royale des Sciences, année 1773, 343-382.

[5]-Croci, G., (2005). “Conservazione e restauro strutturale dei beni architettonici”, Roma

2005.

[6]-Como, M., (2010), “Statica delle costruzioni storiche in muratura. Archi, volte, cupole,

architetture monumentali, edifici sotto carichi verticali e sotto sisma, Roma 2010.

[7]-De La Hire, P., (1730). “Traité de mécanique, où l’on explique tout ce qui est necessaire

dans la pratique des Arts”. Mémoires de l’Académie Royale des Sciences, vol. 9, 1-333.

[8]-De La Hire, P., (1731). “Sur la construction des voûtes dans les édifices”. Histoire de

l’Académie Royale des Science, année 1712, 74-77.

[9]-Di Miceli, (2016), “Restauro della Galleria delle Volte Crollate: l’Ingegneria Strutturale al

servizio della Conservazione dei Monumenti Patrimonio dell’Umanità”, PhD Thesis, in

preparation, Department of Structural Engineering and Geotechniques, Sapienza University of

Rome, Italy.

[10]-D.M.LL.PP. (1986). “Norme tecniche relative alle costruzioni antisismiche”, Decree of

the Italian Ministry of Public Works, G.U. 12-5-1986, n. 108.

[11]-Elia, G.M., (2010). “Metodi e tecniche del restauro architettonico”, Roma 2010.

[12]-Flugge, W., (1973). “Stresses in shells”, Berlino 1973.

[13]-Heyman, J., (1966). “The stone skeleton”. International Journal of Solids and Structures,

2, 249-279.

[14]-Heyman, J., (1982). “The masonry arch”, Ellis Horwood, Chichester 1982.

[15]-Iacopi, I., Tedone, G., (2006). “Bibliotheca e Porticus ad Apollinis”, Mitteilungen des

Deutschen Archaologischen Instituts, Romische Abteilung, 112, 351-378.


-1008-

[16]-LINEE GUIDA (2010). “Linee Guida per la valutazione e riduzione del rischio sismico

del patrimonio culturale allineate alle nuove Norme tecniche per le costruzioni (D.M. 2008)” -

Circolare n. 26/2010.

[17]-Lourenço, P.B., (2002). “Computations on historic masonry structures”, Progress in

Structural Engineering and Materials, 4(3), 301-319, 2002.

[18]-Lugli, G., (1957). “La tecnica edilizia romana: con particolare riguardo a Roma e Lazio”,

Vol. I, Roma 1957.

[19]-Méry, E., 1840. “Sur l’equilibre des voûtes en berceau”. Annales des ponts et Chausséss.

XIX.

[20]-Monti, G., Vailati, M., Gaetani, A., Paolone, A., (2013). “Modelli analitici di capacità

per archi di muratura rinforzati soggetti ad azioni sismiche”, Ingegneria sismica in Italia -

ANIDIS 2013 – XV Convegno, 30 giugno – 4 luglio 2013, Padova, Italy.

[21]-NTC 2008 – DM 14 Gennaio 2008. “Nuove norme tecniche per le costruzioni”. Gazzetta

Ufficiale della Repubblica Italiana, 4 febbraio 2008, n. 29.

[22]-Pensabene, P., Gallocchio, E., (2011). “Elementi architettonici della casa di Augusto sul

Palatino”, Architettura Classica, vol. LXII, n.s. 1, L’Erma di Bretschneider, Roma 2011, 475-

487.

[23]-Pensabene, P., Gallocchio, E., (2013). “Alcuni Interrogativi Sul Complesso Augusteo

Palatino”, Architettura Classica, vol. LXIV, n.s. II, 3, L’Erma di Bretschneider, Roma 2013,

557-582.

[24]-Roca, P., Cervera, M., Gariup, G., Pelà, L., (2010). “Structural analysis of masonry

historical constructions: Classical and advanced approaches”, Archives of Computational

Methods in Engineering, 17(3), 299–325, 2010.

[25]-Tomasoni, E., (2015). “Analisi, verifiche e consolidamento di archi e volte”, Palermo

2015.

Documents

RETROFIT OF A COLLAPSED MASONRY GALLEY WITH …irf/Proceedings_IRF2016/data/... · 2016-05-27 · Since the XV century studies about the static vaults were developed, but the first